RSV Prev CE Analysis

Pharmacoeconomics 2007; 25 (1): 55-71
ORIGINAL RESEARCH ARTICLE 1170-7690/07/0001-0055/$44.95/0
© 2007 Adis Data Information BV. All rights reserved.
Cost Effectiveness of Palivizumab for
Respiratory Syncytial Virus
Prophylaxis in High-Risk Children
A UK Analysis
Mark J.C. Nuijten,1,2 Wolfgang Wittenberg3 and Maximilian Lebmeier3,4
1 Erasmus University, Amsterdam, The Netherlands
2 MEDTAP, Amsterdam, The Netherlands
3 Abbott GmbH & Co. KG, Ludwigshafen, Germany
4 University of Sheffield, Sheffield, England
Objective: To assess the cost effectiveness of palivizumab (a preventative treat-Abstract
ment against severe respiratory syncytial virus [RSV] infection) in children at
high risk of hospitalisation, i.e. preterm infants ≤35 weeks gestation, children with
bronchopulmonary dysplasia (BPD) and children with congenital heart disease
(CHD).
Methods: A decision tree model was developed employing data sources from the
published literature, palivizumab clinical trials, official UK price/tariff lists and
national population statistics. The comparator was no prophylaxis. The primary
perspective of the study was that of the UK NHS. In a societal perspective
scenario analysis, the future lost productivity of a child resulting from RSV-
related mortality (indirect costs) was also included. The cost of administration of
palivizumab, hospital care for RSV infections and the cost of asthma treatment
were included.
The analysis was based on a lifetime follow-up period in order to capture the
impact of palivizumab on long-term morbidity and mortality resulting from an
RSV infection. The primary efficacy outcome in the palivizumab clinical trials
was the number of RSV hospitalisations avoided, which was extrapolated to
effectiveness outcomes, i.e. number of life-years gained and number of QALYs.
Costs and effects were discounted by 3.5%.
Results: In preterm infants and children with BPD, prophylaxis with palivizumab
compared with no prophylaxis had an incremental cost-effectiveness ratio (ICER)
of £7042/QALY without discounting outcomes, increasing to £16 720/QALY
after discounting. In babies with CHD, the use of palivizumab resulted in an ICER
of £2427/QALY without discounting outcomes and £6664/QALY after discount-
ing. One-way sensitivity analyses and probabilistic sensitivity analyses confirmed

56 Nuijten et al.
the robustness of the model. A scenario analysis showed that the inclusion of
indirect costs leads to further improvement in the cost-effectiveness outcomes for
palivizumab.
Conclusion: This study suggests that palivizumab prophylaxis against severe
RSV infection in children at high risk may be cost effective from the NHS
perspective (vs no prophylaxis), and that the positive clinical and economic
benefits may persist beyond one RSV season.
Respiratory syncytial virus (RSV) is a common all mortality rate in these patients was also signifi-
virus that is not restricted by geographical, cultural cantly higher than the control cohort (8.11% vs
or socioeconomic factors. The incidence of RSV 1.58%; p = 0.001).[6]
infection in children is high; by the age of 1 year There is evidence that early RSV infection pre-
25–50% will have been infected and by the age of 5 disposes children to recurrent wheezing during their
years 95% will have been infected.[1] RSV is a early childhood, but that airway morbidity is tran-
leading cause of lower respiratory tract illness in sient and subsides during school age.[7] Among
children, the elderly and immunocompromised indi- young school children, previous RSV infection may
viduals. Preterm infants, children with bronchopul- increase the risk of asthma 10-fold.[8]
monary dysplasia (BPD) and children with congeni-
As current treatment options for severe RSV
tal heart disease (CHD) are particularly at risk of
infection are limited to supportive care, the focus
hospitalisation due to severe RSV infection.[2] Hos-
has shifted to prevention. Two products have been
pitalisation rates in these at-risk patients may be as
developed successfully: RSV intravenous immu-high as 20%.[3]
A UK study reported a national
noglobulin (RSV-IGIV) and palivizumab. Only thehospital admission rate of 3% for RSV infection.[2]
latter is currently marketed in the UK. PalivizumabAnother UK study by the Public Health Laboratory
is given as an intramuscular injection. The recom-Service reported that RSV infection was the most
mended dose is 15 mg/kg bodyweight, given once acommon cause of hospital admissions due to acute
month during anticipated periods of RSV circulationrespiratory illness in young children in the UK.[4]
A
in the community.[9,10]
study by Clark et al.[5]
in the UK assessed the rate of
Palivizumab was initially licensed for use in chil-RSV hospitalisation in unprophylaxed children with
dren at a high risk of RSV infection, i.e. childrenBPD in Liverpool over two RSV seasons (1998/99
born at ≤35 weeks of gestation and who are <6and 1999/2000) to be 19.7%.
months old at the onset of the RSV season andInvestigations of the long-term prognosis of pa-
children <2 years old who have received treatmenttients with RSV disease in infancy have shown
for BPD within the proceeding 6 months. IMpactmeasurable respiratory abnormalities immediately
(International Multicenter Phase III Clinical Trial inor several years following infection. Sampalis[6]
in-
RSV), a large multinational clinical trial, has report-vestigated the morbidity and mortality after RSV-
ed a significant reduction (relative reduction [RR]associated hospitalisations among preterm Canadian
55%; 95% CI 38, 72) in the overall rate of hospital-infants. The odds ratios calculated showed that in-
isations due to RSV infection in these high-riskfants with an RSV hospitalisation were at least 8-
children given palivizumab compared with thosefold more likely to be hospitalised for respiratory
conditions during the follow-up period and the over- given placebo[10] (4.8% vs 10.6%; p = 0.0004).
© 2007 Adis Data Information BV. All rights reserved. Pharmacoeconomics 2007; 25 (1)

Cost Effectiveness of Palivizumab for RSV Prophylaxis 57
No sequelae
Sequelae
Stay alive
Die
RSV-hospitalisation,
including ICU
No RSV-hospitalisation
Palivizumab
No sequelae
Sequelae
Stay alive
Die
including ICU
No prophylaxis
Preterm
BPD
[+]
[+]
Preterm/BPD
No sequelae
Sequelae
Stay alive
Die
including ICU
Palivizumab
No sequelae
Sequelae
Stay alive
Die
including ICU
No prophylaxis
Cyanotic
Acyanotic
CHD
High-risk
children
Fig. 1. Model structure: treatment pathway for children with respiratory syncytial virus (RSV) infection. BPD = bronchopulmonary dysplasia;
CHD = congenital heart disease; ICU = intensive care unit.
The licence for palivizumab was later extended Methods
for use in children with haemodynamically signifi-
cant CHD. Feltes et al.[11]
showed that, in children
Model Design and Study Populations
with CHD, palivizumab recipients had a 45% RR
(95% CI 23, 67) in hospitalisations due to RSV A decision tree model was developed employing
infection compared with those receiving placebo data sources from the published literature,
palivizumab clinical trials, official UK price/tariff(5.3% vs 9.7%; p = 0.003). Adverse events were
lists and national population statistics.similar in the treatment groups and no child discon-
A local external health economics expert validat-tinued the drug for a related adverse event.[11]
ed the research plan for the model (structure, as-
This study examines whether, from the perspec-
sumptions and data sources) taking into considera-
tive of the UK NHS, prevention of RSV infection
tion National Institute of Health and Clinical Excel-
with palivizumab is a cost-effective strategy in lence (NICE) guidance and the specific UK
preterm infants or children with BPD or CHD. In an healthcare setting. The final model was subsequent-
analysis from the societal perspective, the future lost ly approved by the external expert.
productivity of a child resulting from RSV-related The populations in the model reflected the inclu-
mortality (indirect costs) is also included. sion and exclusion criteria of the IMpact[10] (preterm

58 Nuijten et al.
Table I. Clinical data and sources: primary data for the base model
Parameter Probability p-Value Reference
palivizumab no prophylaxis
Prophylaxis in preterm infants and children with BPD
Hospitalisation
overall 0.048 0.106 <0.001 10
preterm, excluding BPD 0.018 0.081 <0.001 10
BPD 0.079 0.128 0.038 10
Prophylaxis in children with CHD
Hospitalisation
overall 0.053 0.097 0.003 11
cyanotic 0.056 0.079 NS 11
acyanotic 0.05 0.118 0.003 11
Mortality
trial 0.033 0.042 NS 11
BPD = bronchopulmonary dysplasia; CHD = congenital heart disease; NS = not significant.
infants born ≤35 weeks of gestation or children with palivizumab on long-term morbidity and mortality
BPD) and Feltes et al.[11]
(children with CHD) trials. resulting from an RSV infection. The effect of a
A logistic regression analysis reported for IMpact reduction in RSV infections was limited to prophy-
showed that “gestational age was not a significant laxis during a single RSV season, corresponding to
predictor of RSV hospitalisation.” Therefore, the data available from the palivizumab clinical trial
various gestational subgroups were considered as (IMpact: children were followed for 150 days from
one ‘preterm’ population in the model. Children in the point of random assignment). Morbidity result-
the ‘CHD’ population were stratified into two sub- ing from RSV infection consisted of asthma and its
populations: ‘cyanotic’ and ‘acyanotic’, correspond- associated utilities and costs.
ing with the stratification in the clinical trial.[11]
Perspective and Setting
Figure 1 shows the structure of the model for the
The perspective of the study in the base-case
‘preterm/BPD’ and ‘CHD’ populations. Children
analysis was that of the UK NHS as recommended
may develop an RSV infection leading to hospital-
by NICE guidance for manufacturers and spon-
isation. The majority of these children will be man-
sors.[12] The base-case analysis did not include costs
aged in a paediatric ward, but a proportion will
that are relevant from the broader ‘societal’ perspec-
require transfer to a neonatal or paediatric intensive
tive, e.g. ‘indirect costs’ relating to the future pro-
care unit (NICU or PICU). A small proportion of the
ductivity losses of the child or the parents.
children may die. A significant proportion of the
In the societal perspective scenario analysis, the
surviving children may develop long-term sequelae
future lost productivity of a child resulting from
(e.g. wheezing, asthma).[6,9] These clinical assump-
RSV-related mortality (indirect costs) was also in-
tions, and others on which the structure of the model
cluded.
was based, were supported by data from the IM-
Clinical Outcomespact[10] and Feltes et al.[11] trials.
The primary efficacy outcome in the palivizumab
Time Horizon clinical trials was the number of RSV hospitalisa-
The base-case analysis was based on a lifetime tions avoided.[10,11]
In the model, the efficacy data
follow-up period in order to capture the impact of were extrapolated to effectiveness outcomes: num-

ber of life-years gained (LYG) and number of the UK setting. Clinical events may be considered as
QALYs. In the base-case analysis, clinical outcomes not being country specific[14] and therefore can be
were discounted at a rate of 3.5% in line with current derived from international studies. For economic
NICE guidance. measures and information on therapeutic choices,
country-specific data sources should be used.[14]
Cost Assessment
An international literature search was performed
The cost assessment in the base-case analysis
on all relevant input data for the model: probabilities
was based on the assignment of cost estimates to the
of hospitalisation and mortality, life expectancy,
following:
utilities, resource utilisation, costs and asthma. The
• Palivizumab prophylaxis for RSV infection, con-
search on these terms was constrained by the search
sisting mainly of the cost of palivizumab and
term ‘RSV’ and included the following terms: RSV,
resource utilisation related to the administration
costs, mortality, morbidity, asthma, QALY, utility.
of palivizumab. No costs were assigned to the ‘no
The results are described below.
prophylaxis’ arm of the model.
Table I shows the primary input data for the
• Hospitalisation due to RSV infection.
model: clinical trial data for hospitalisations, clinical
• The clinical complications of RSV infection, i.e. data on wheezing and observational data on mortali-
cost of asthma treatment. ty and life expectancy.
Economic outcomes were discounted at a rate of
Respiratory Syncytial Virus (RSV) Hospitalisation3.5% when the time horizon of the model extended
beyond 1 year. The model cost inputs are taken from Prophylaxis in Preterm Infants and Children with
Bronchopulmonary Dysplasia (BPD)different years because of the availability of data.
All have been corrected to 2003 values as no 2004 The base-case analysis was based on the IMpact
inflation data were available at the time of this trial.[10]
Scenario analyses were performed on sub-
study.[13]
populations because of differences observed in hos-
pitalisation rates between treatment groups for theCosts of adverse events were not considered be-
subgroups in the IMpact trial; however, the trial wascause the clinical trials did not show any difference
not powered at the subpopulation level. A review ofbetween prophylaxis and no prophylaxis.[10,11]
data from the clinical trials[15,16]
comparing RSV-
Data Sources IGIV prophylaxis with no prophylaxis showed that
the hospitalisation rates for no prophylaxis were
Various data sources were considered for the much higher (13.5% and 20.2%; see table I of the
model in order to maximise its external validity for supplementary material [“ArticlePlus”] at www.
adisonline.com/phe) than in the IMpact trial
(10.6%).[10]
As this indicates that the IMpact trial
may have underestimated the actual risk of hospital-
isation due to RSV, we included a scenario analysis
using adjusted values based on these other data.[15,16]
An indirect assessment of the hospitalisation risk
for palivizumab was made by applying the RR for
hospitalisation in patients receiving palivizumab
versus no prophylaxis (4.8%/10.6%) from the IM-
pact trial to the hospitalisation rate of the ‘no pro-
Table II. Resource utilisation for respiratory morbidity[29]
Resource utilisation RSV Control
Days in general ward 39.8 2.2
Days in ICU 2.9 0
Days in paediatric ward 30.7 1.6
Other (general ward; days) 6.2 0.6
Outpatient attendances 11.9 8.6
GP contacts 16.3 14.6
Community care contacts 28.2 18.7
Consultations for GP for respiratory illness 8.3 5.6
ICU = intensive care unit; RSV = respiratory syncytial virus.

60 Nuijten et al.
nario analysis was performed in order to capture the
uncertainty of the actual risk of hospitalisation: the
probability for a hospitalisation in palivizumab-
treated patients was adjusted to 8.2% by applying
the RR from the Feltes et al.[11] trial (5.3%/9.7%) to
the ‘no prophylaxis’ hospitalisation rate (15.0%) in
the Cardiac study.[17]
Mortality
Clinical studies evaluating the overall benefit of
Table III. Unit costs (£, 2003 values) employed in the model
Item Cost/unit Reference
All patients
per diem paediatric ward 436 34
per diem in paediatric ICU 1526 34
Long-term only
per diem in general ward 210 34
outpatient paediatric 131 34
attendance
GP contact 20 35
community care contacts 27 29
ICU = intensive care unit.
prophylaxis against RSV infections have focused on
the prevention of RSV-related hospitalisations as
phylaxis’ arm of the Groothuis et al.[15]
and PRE-
the single outcome measure. Most clinical studies,
VENT[16]
(PREVENTion of respiratory syncytial
including the IMpact[10] and Feltes et al.[11] studies,
virus) studies:
also show that prevention of RSV leads to an abso-
• Groothuis et al.: the adjusted probability for lute reduction in mortality. Reported mortality rates
palivizumab is 9.2% (20.2% × 4.8%/10.6%). were 1% and 0.4% for no prophylaxis and
• PREVENT: the adjusted probability for palivizumab, respectively, in the IMpact[10] study
palivizumab is 6.1% (13.5% × 4.8%/10.6%). and 4.2% and 3.3% in the Feltes et al.[11]
study.
These studies were not powered to yield statisticallyProphylaxis in Congenital Heart Disease (CHD)
significant differences in mortality, although the
The stratification of the data,[10] which were not
absolute mortality was lower for palivizumab. How-
prespecified, did not yield statistically significant
ever, evidence exists in the literature that an RSV
differences between cyanotic patients receiving
infection in high-risk children is associated with
palivizumab prophylaxis or no prophylaxis, because
significantly higher mortality risks than in non-in-
the trial was not powered at the subpopulation level.
fected children.
Therefore, the base-case analysis used the overall
Preterm Infants and Children with BPDprobability for hospitalisation. A scenario analysis
was based on the specific probabilities for cyanotic An observational study by Sampalis[6] showed a
and acyanotic patients. The study by the Cardiac statistically significant increase in mortality follow-
Study Group[17] yielded a much higher hospitalisa- ing hospitalisation due to RSV infection. In this
tion rate (15.0%) for children with CHD in the study, data were analysed from 2415 preterm infants
control group than the Feltes et al.[11]
study (9.7%). (32–35 weeks gestational age) hospitalised for prov-
The Cardiac study examined the effectiveness of en or probable RSV and 20 254 matched control
RSV-IGIV in reducing hospitalisation for treatment infants. The patients in both cohorts were matched
of RSV in children with CHD. Children aged <4 with respect to potential confounders. The exclusion
years of age were randomly assigned to a treatment of infants with BPD and congenital abnormalities in
group receiving RSV-IGIV or to a control group not the control group also removed the possibility of
receiving infusions. Therefore, the hospitalisation confounding. However, the mortality rate for the
rates of the control group of this study can be RSV cohort is probably an underestimation of the
compared with hospitalisation rates in the placebo mortality for children with BPD or preterm infants
group in the Feltes et al.[11]
study and should reflect born earlier than 32 weeks. The databases used
the natural incidence of RSV hospitalisation. A sce- contained data on all births in a hospital and for all

hospitalisations in the target population. The overall surgery have improved mortality rates from RSV
mortality rate in the RSV cohort in the 2 years after infection in this high-risk population. Various stud-
the initial hospitalisation was 8.11% compared with ies have reported mortality rates varying between
1.58% in the control cohort (p = 0.001), showing a 3% and 5%.[20,21] A limitation of most of these
significant risk of mortality beyond the initial hos- studies was that, in addition to children with CHD,
pitalisation for RSV. This was used in the base-case they also included preterm infants and children with
analysis. chronic lung disease, who have a significantly lower
In contrast, Joffe et al.[18]
estimated that the mortality risk than children with CHD. Therefore,
probability of death among high-risk infants who these studies may have underestimated the mortality
were hospitalised for RSV infection was 1.2% (95% risk of children with CHD.
CI 0, 2.8). This figure was derived by pooling place- In order to determine the mortality in CHD, data
bo and intervention subjects from three trials
from a meta-analysis by Wang and Tang[22]
were
(Groothuis et al.,[15] PREVENT[16] and IMpact[10]).
used. This meta-analysis determined the hospitalisa-
We used the more conservative values from Joffe et
tion and mortality rates for the total high-risk popu-
al.[18] for the sensitivity analyses, as these probabili-
lation.[3,23]
This population also included preterm
ties only reflect mortality during the hospitalisation
infants and children with BPD. Therefore, we per-
period and do not capture the long-term mortality
formed an analysis including only children with
resulting from an RSV infection.
CHD, and estimated that the mortality rate for the
CHD population who were hospitalised for RSVChildren with CHD
was 4.5%, which is significantly higher than theA prospective study conducted in the early 1980s
0.92% reported mortality rate for children withoutshowed a high mortality rate of 37% for RSV-
CHD. From this analysis the ratio of mortality toinfected, hospitalised children with CHD, with
hospitalisations (P_mort/hos) in children with CHDhigher rates in children with nosocomial infections
was 0.40 (mortality rate of 4.5% and hospitalisationand/or pulmonary hypertension.[19] Advances in
rate of 11.2%). This calculation was based on pool-ICU and reparative and palliative congenital heart
Table IV. Base-case analysis results for prophylaxis in preterm infants and children with bronchopulmonary dysplasia: costs (£, 2003
values) and outcomesa
Parameter Palivizumab No prophylaxis
Costs
Direct medical 4281 521
Asthma/recurrent wheezing 668 1476
Direct medical (including asthma/recurrent wheezing) [base-case] 4949 1996
Indirect 764 1687
Direct medical + indirect 5044 2207
Direct medical (with asthma/recurrent wheezing) + indirect 5712 3683
Outcomes
No discounting
life-years lost 0.30 0.66
QALYs 75.95 75.53
Discounting
QALYs 26.71 26.53
a Costs and outcomes were discounted at 3.5% unless otherwise indicated.

62 Nuijten et al.
Table V. Scenario analyses for preterm infants and children with bronchopulmonary dysplasia (BPD). Costs (£) are in 2003 valuesa
Outcome No discounting Discounting
difference difference ICER difference difference ICER
in costs in outcomes in costs in outcomes
Preterm
LYG 2 858 0.39 7 256 2 858 0.14 20 344
QALY 2 858 0.46 6 272 2 858 0.19 14 883
BPD
LYG 3 122 0.31 10 189 3 122 0.11 28 569
QALY 3 122 0.35 8 819 3 122 0.15 20 953
Scenario analysesb
Groothuis[15]
LYG 1 961 0.69 2 834 1 961 0.25 7 945
QALY 1 961 0.80 2 455 1 961 0.34 5 839
PREVENT[16]
LYG 2 658 0.46 5 770 2 658 0.16 16 179
QALY 2 658 0.53 4 994 2 658 0.22 11 862
a Costs and outcomes were discounted at 3.5%.
b Using hospitalisation rates from alternative data.
ICER = incremental cost-effectiveness ratio; LYG = life-years gained; PREVENT = PREVENTion of respiratory syncytial virus study.
ing the data for the CHD children. Weighting by national data on population statistics. The life ex-
means of the general variance method yielded simi- pectancy of healthy children is 78.05 years and
lar results: hospitalisation rate of 10.8%, mortality therefore the life expectancy at age 1 year was
of 4.3% and P_mort/hos of 39.4%. In order to calcu- assumed to be approximately 77.05 years for these
late the mortality rate, this value was multiplied by children.[24]
the rate of RSV hospitalisations from Feltes et al.[11]
Children with CHD
Using this method to calculate mortality rates for
In a UK study by Wren and O’Sullivan[25]
of
children given prophylactic palivizumab and those
1942 cases of CHD diagnosed in infancy in a popu-
given no prophylaxis, generated similar values to
lation of 377 310 live births (5.2/1000), 1588 (82%)
those reported in the study:[11]
survived to 1 year and 1514 (78%) were predicted to
• palivizumab: 2.3% (calculated) versus 3.3% (re-
survive to age 16 years. Thus, the value used in the
ported).
model (the proportion surviving from year 1 to year
• no prophylaxis: 4.0% (calculated) vs 4.2% (re-
16) was 95.3% (annual mortality 0.32%). It was
ported) [see also the supplementary material
assumed that beyond 16 years the children would
(“ArticlePlus”) at www.adisonline.com/phe].
have a normal life expectancy of 62.05 years
(78.057 – 16.0).[24]
Life Expectancy
Preterm Infants and Children with BPD Utilities
There were no clinical data on lower life-ex- A study by Greenough et al.[26] reported utilities
pectancy subsequent to RSV hospitalisation for in- in children with a history of RSV hospitalisation.
fants born preterm and children with BPD. There- The parents were asked to complete 15 questions
fore, it was assumed that the life expectancy for about their child’s health over the previous 4 weeks.
these patients was similar to the life expectancy for The median Health Utility Index (HUI)-2 multi-
children at 1 year of age in the UK, derived from attribute utility function was 0.88 (range 0.16–1.00)

in the RSV-proven children, which was significantly sponding with a median HUI-2 for non-RSV
lower (p = 0.0088) than that of the non-RSV chil- infected children (u = 0.95).
dren (median 0.95, range 0.03–1.00). The median • High-risk children who experienced a hospital-
HUI-3 multi-attribute score of the two groups did isation due to a RSV infection would have a
not differ significantly (0.93, range 0.05–1.00 vs utility corresponding with a median HUI-2 for
0.97, range 0.32–1.00). The following assumptions RSV-infected children (u = 0.88).
were made for the model base-case analysis: • Beyond the age of 16 years there would be no
• High-risk children who did not experience a hos- difference in utility between the various high-risk
pitalisation due to RSV infection would not have patients regardless of the development of RSV
perfect health, but would have a utility (u) corre- infection and the development of long-term res-
Table VI. Sensitivity analyses (direct medical costs [£, 2003 values], including asthma) for preterm infants and children with bronchopulmo-
nary dysplasia
Parameter Outcome No discounting Discounting
LOS
Thomas et al.[28]
LYG 2 950 0.36 8 132 2 950 0.13 22 802
QALY 2 950 0.42 7 034 2 950 0.18 16 702
Netten and Curtis[27]
LYG 2 950 0.36 8 133 2 950 0.13 22 803
QALY 2 950 0.42 7 035 2 950 0.18 16 703
Vials
10% reduction LYG 2 549 0.36 7 028 2 549 0.13 19 705
QALY 2 549 0.42 6 079 2 549 0.18 14 434
20% reduction LYG 2 143 0.36 5 909 2 143 0.13 16 568
QALY 2 143 0.42 5 111 2 143 0.18 12 136
Thomas et al.[28]
LYG 973 0.36 2 682 973 0.13 7 519
QALY 973 0.42 2 320 973 0.18 5 508
Utilities – 0.93[26]
LYG 2 953 0.36 8 141 2 953 0.13 22 826
QALY 2 953 0.36 8 253 2 953 0.13 23 578
Incidence of recurrent wheezing[31]
LYG 2 953 0.36 8 141 2 953 0.13 22 826
QALY 2 953 0.49 6 065 2 953 0.22 13 031
Administration-consult (assumption) LYG 3 591 0.36 9 900 3 591 0.13 27 758
QALY 3 591 0.42 8 563 3 591 0.18 20 332
Mortality
trial LYG 2 953 0.46 6 387 2 953 0.16 18 226
QALY 2 953 0.52 5 702 2 953 0.21 13 981
palivizumab = 0 LYG 2 953 0.77 3 832 2 953 0.27 10 745
QALY 2 953 0.82 3 589 2 953 0.32 9 272
palivizumab = placebo LYG 2 953 0.00 a 2 953 0.00 a
QALY 2 953 0.06 48 973 2 953 0.05 58 685
Discounting LYG 2 938 0.36 8 106 2 963b 0.08b 36 034b
QALY 2 938 0.42 6 949 2 963b 0.12b 24 157b
a No prevention dominant.
b Discounting costs and outcomes at 6%.
ICER = incremental cost-effectiveness ratio; LOS = length of stay; LYG = life-years gained.

64 Nuijten et al.
each infant would receive 4.87 monthly doses of
palivizumab over the winter/spring season, corre-
sponding with children studied in the IMpact study
(4.93 doses in Feltes et al.[11]).[10]
For the base-case analysis, the administration
(drug utilisation) of palivizumab was based on as-
sumptions about the number of 50mg and 100mg
vials used in the palivizumab arm of the trials. We
assumed that:
• 38.7% of the preterm babies and the children
with BPD used a 50mg vial and 91.3% used a
100mg vial; these assumptions are based on the
reported weight of the infants.
• 39.6% of the children with CHD used a 50mg
vial and 100% used a 100mg vial; in 3.8% of
0
10 000
20 000
30 000
40 000
50 000
60 000
70 000
80 000
90 000
100 000
0 2 4 6 8 10
Mortality (%)
£/QALY
Medical costs
Total costs
Fig. 2. Sensitivity of the base-case incremental cost-effectiveness
ratio (2003 values) to the effect of using mortality rates from
Sampalis.[6] We used a range from 0 to 10% to assess the robust-
ness of the model to mortality, which includes the 8.11% mortality
from the Sampalis study.
cases a second 100mg vial was assumed to have
been administered. These assumptions are basedpiratory morbidity following a hospitalisation
on the average reported weight of 6.050kg fordue to RSV infection.
infants in the Feltes et al. trial.[11]
• All patients aged >16 years would have perfect
Thomas et al.[28] used one 100mg vial per infant;health (u = 1).
a scenario analysis was conducted based on thisAs the median HUI-3 multi-attribute score of the
administration regimen.two groups did not differ significantly, a scenario
It was assumed that each administration was per-analysis was conducted based on a utility of 0.93 (no
formed during a routine paediatric follow-up visitRSV hospitalisation).
and did not incur additional costs. A sensitivity
Resource Utilisation and Costs
analysis was performed to include an administration
Table II shows the resource utilisation based on a
cost based on the assumption that each administra-
literature review and an analysis of the clinical tri-
tion would require a visit to a hospital outpatient
als.[27-29] Due to the need for country-specific data
facility, which includes a specialist fee.
for resource utilisation, the literature review focused
on the UK and included search terms related to RSV
Days of RSV Hospitalisation in Preterm Infants and
and its sequelae (e.g. asthma, mortality). These UK Children with BPD
data on resource utilisation were not stratified ac-
A UK study by Netten and Curtis[27]
reported an
cording to the indications in this model (preterm
average length of stay (LOS) of 4 days for non-
infants without BPD and children with BPD and
intensive paediatric care and 4 days for intensive
CHD).[29]
care for RSV hospitalisation.[30] Thomas et al.[28]
Units of Palivizumab reported an average LOS of 3.2 days for the paedia-
Palivizumab prophylaxis consisted of a dose of tric ward and 4.3 days for the ICU. As no underlying
15 mg/kg injected intramusculary once a month for data on the patient population for the Netten and
4 or 5 months (93% received all planned five injec- Curtis calculation were available and the Thomas et
tions). In the base-case analysis it was assumed that al.[28] study was based on a small sample size (n = 6),

the LOS data from the IMpact trial were consid- ducted based on data from the studies by Netten and
Curtis[27] and Thomas et al.[28] in order to test theered.[10]
sensitivity of the selection of the trial data.An analysis of the underlying IMpact trial data
was performed by pooling the data on LOS for RSV
Days of RSV Hospitalisation in CHDhospitalisations from both study arms. It was justifi-
able to pool the data from both arms, as the length of No specific data regarding LOS were found for
stay only related to the RSV infection, and is not RSV infections in CHD. Therefore, an analysis of
related to the use of palivizumab or no prophylaxis. the underlying Feltes et al.[11] trial data was per-
This is a prudent approach, because the IMpact formed by pooling the data on LOS of RSV hospital-
trial[10]
showed an absolute significant reduction in isations from both study arms. Again it was assumed
probability of ICU admission for infants who re- that the LOS relates only to the RSV infection and
ceived palivizumab (vs placebo), which suggests a not to the use of palivizumab or no prophylaxis.
positive impact on the severity of the RSV infection, Although there was an absolute reduction in
and consequently shorter LOS. The total average probability of ICU admission for infants who re-
LOS was 6.45 days (4.58 days for the paediatric ceived palivizumab (vs placebo), the difference was
ward and 1.88 days for the ICU). An analysis of a not statistically significant. The total average LOS
subset of UK patient data from IMpact[10]
showed was 12.03 days: (6.40 for the paediatric ward and
that average LOS was 7.84 days (6.47 days for the 5.63 for the ICU). An analysis of a subset of UK
paediatric ward and 1.37 days for the ICU), but this data showed that the average LOS was 12.39 days
subset included data from only seven hospitalisa- (6.25 for the paediatric ward and 6.14 for the ICU).
tions. As the average LOS from the complete IM- As the average LOS from the complete trial data set
pact[10] data set was similar to that of the UK subset was similar to that of the UK subset (12.03 and
(6.45 and 7.84 days, respectively), it was decided 12.39 days, respectively), it was decided that the
that the mean UK-specific data would be used from mean UK-specific data would be used for the base-
the IMpact trial,[10]
and that the total data set would case analysis. For the probabilistic sensitivity analy-
be used to create a distribution for LOS, which was sis the distribution of LOS between the ICU and
applied to the UK-specific mean for the probabilistic paediatric wards from the total data set was applied
to the UK-specific mean.sensitivity analysis. A scenario analysis was con-
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Cost-effectiveness threshold (£ × 1000)
Probabilityofbeingcosteffective
No prevention
Palivizumab
Fig. 3. Cost-effectiveness acceptability curve for palivizumab prophylaxis in preterm infants and children with bronchopulmonary dysplasia
(discounting effectiveness and costs; 2003 values).

66 Nuijten et al.
tial future lost productivity resulting from RSV-
related mortality. The unit cost of 1 hour of produc-
tive time was £8.53 (1998 values) for manufacturing
labour,[32] and this was inflated to £10.15 per hour
using an inflation correction factor from national
statistics. Data from the Organisation for Economic
Cooperation and Development (OECD) was used to
estimate the total time worked annually by an em-
ployed person (1673 hours) and the employment
rate (0.729).[33]
These estimates led to an annual
productivity cost estimate of £12 382.
−5000
−4000
−3000
−2000
−1000
0
1000
2000
3000
4000
5000
−0.4 −0.3 −0.2 −0.1 0 0.1 0.2 0.3 0.4
QALYS
Costs
£25 000/QALY
Fig. 4. Incremental cost-effectiveness plane for palivizumab pro-
phylaxis in preterm infants and children with bronchopulmonary
dysplasia (discounting effectiveness and costs; £, 2003 values).
Results
Cost of Asthma and Recurrent Wheezing
Greenough et al.[29]
compared the use of health-
care resources between children with CHD who had Prophylaxis in Preterm Infants and Children
or did not have an admission with a proven RSV with BPD
infection in the UK, and showed that an RSV infec-
The base-case analysis was based on medicaltion increases medical resource use. The resource
costs, including the cost of asthma treatment (tableutilisation from this study (table II) was used for the
IV). The use of palivizumab resulted in an incre-model and applied to all patient populations. The
mental cost-effectiveness ratio (ICER) of £7042/total costs over 2 years were calculated by multiply-
QALY without discounting of effectiveness anding the units of resource utilisation by the identified
£16 720/QALY after discounting. The cost/LYGcosts (table III). The Greenough et al.[29] study in-
was slightly higher: £22 826 and £8141, respective-cluded the initial hospitalisation due to RSV infec-
ly, with and without discounting effectiveness.tion, and therefore an adjustment was made by sub-
tracting the costs for the first hospitalisation, con-
tained in the present model, from the 2-year total
costs in the study by Greenough et al.[29] Thus, all
costs of sequelae beyond 2 years from the initial
RSV hospitalisation were not included in the model
because of lack of longer term data, which can be
considered a conservative approach for palivizum-
ab.
A scenario analysis was performed using the
study by Simoes et al.,[31] which shows that
palivizumab use decreases the risk of recurrent
wheezing in preterm infants from 19.1% to 6.8%.
Indirect Costs (Productivity Losses)
The scenario analysis, which was performed
from a societal perspective, also included the poten-
Table VII. Base-case results for prophylaxis in congenital heart
disease: costs (£, 2003 values) and outcomes
Parameter Palivizumab No prophylaxis
Costs
Direct medical 5143 1173
Asthma 738 1350
Direct medical (including asthma) 5881 2523
Indirect 4071 7451
Direct medical + indirect 9214 8624
Direct medical (with asthma) 9952 9974
+ indirect
Outcomes
No discounting
QALYs 72.32 70.94
Discounting
QALYs 25.29 24.79

Table VIII. Scenario analyses for prophylaxis in children with congenital heart disease. Costs (£) are in 2003 values
Outcome No discounting Discounting (3.5%)
Cyanotic[11]
LYG 3 904 0.71 5 523 3 904 0.25 15 575
QALY 3 904 0.72 5 398 3 904 0.26 14 816
Acyanotic[11]
LYG 2 733 2.09 1 308 2 733 0.74 3 688
QALY 2 733 2.14 1 279 2 733 0.78 3 512
Scenario analyses
Cardiac[17]
LYG 2 732 2.09 1 307 2 732 0.74 3 685
QALY 2 732 2.14 1 279 2 732 0.78 3 516
ICER = incremental cost-effectiveness ratio; LYG = life-years gained.
Table IV also shows the impact of the inclusion when the actual mortality data from the IMpact
of indirect costs. The ICER for palivizumab was trial[10]
were used. A sensitivity analysis that varied
£4841/QALY without discounting effectiveness and the mortality for palivizumab from zero to the mor-
£11 494/QALY with discounting. tality rate for no prophylaxis showed outcomes for
cost effectiveness within the generally acceptedTable V shows the results of the subpopulation
range (figure 2).and scenario analyses. The ICERs for the preterm
Figure 3 shows the cost-effectiveness acceptabil-and BPD subpopulations were £14 883 and £20 953
ity curve and figure 4 shows the incremental cost-per QALY, respectively, after discounting. The sce-
effectiveness plane. These figures were based on anario analysis showed that the results were sensitive
probabilistic sensitivity analysis with effectivenessto the rate of hospitalisation. The ICER according to
discounted. This analysis, and that for the CHDthe Groothuis et al.[15]
scenario was substantially
population, was based on beta distributions for hos-lower than the base-case results (£5839/QALY vs
pitalisation probabilities, mortality and vial use, and£16 720/QALY). The PREVENT[16]
scenario also
on a Normal and log-Normal distribution for LOSlead to lower results (£11 862/QALY vs £16 720/
and resource use for asthma, respectively. This anal-QALY).
ysis did not include indirect costs (productivityTable VI summarises the results of various other
losses). The cost-effectiveness acceptability curvesensitivity analyses, which show that the ICER was
shows the probability was 0.73 that the ICER ofsensitive to discounting. The main reason is that
palivizumab was less than £25 000/QALY and 0.86discounting has no effect on the short-term costs.
that the ICER was less than £30 000/QALY.The ICER without discounting of costs and effec-
tiveness was £6949/QALY. Discounting outcomes
at 6% yielded an ICER of £24 157/QALY. The Prophylaxis in CHD
outcome of the model was also very sensitive to the
vial use. This sensitivity analysis shows that the The use of palivizumab resulted in an ICER of
base-case analysis for palivizumab was based on a £2427/QALY without discounting for effectiveness
conservative approach. and £6664/QALY after discounting for infants with
The sensitivity analyses for mortality show that CHD. The cost/LYG was slightly higher: £2483 and
cost effectiveness remained approximately similar £7002 before and after discounting effectiveness,

68 Nuijten et al.
respectively. Table VII also shows that palivizumab Discussion
was dominant after the inclusion of indirect costs.
This study used a decision analytic tree to ex-Table VIII shows that the cost effectiveness of
amine the cost effectiveness of palivizumab inpalivizumab varies substantially between the two
preterm infants, children with BPD and childrenCHD subpopulations: the ICER with discounting for
with CHD. The model was sensitive to the discount-
effectiveness in the model for the ‘cyanotic’ sub-
ing of effectiveness because discounting has no ef-
population is £14 816/QALY and that for the ‘acya-
fect on the short-term costs but affects the long-term
notic’ subpopulation is £3512/QALY. effectiveness outcomes (LYGs and QALYs), thus
The hospitalisation rate for the ‘no prophylaxis’ the ICER results are much higher after discounting.
Nevertheless, the results remain within the range ofarm in the IMpact[10]
study was lower than in the
what is considered acceptable when using a thresh-Cardiac study.[17] A scenario analysis shows that the
old of £25 000/QALY.[36]
results of the model were sensitive to these differ-
A scenario analysis showed that the exclusion ofences in hospitalisation rates. Using data from the
asthma costs only leads to a moderately higher
Cardiac study, the ICER without discounting for
ICER for palivizumab prophylaxis for all patient
effectiveness was £3516/QALY (table VIII) versus
populations. Inclusion of indirect costs leads to a
£6664/QALY in the base case (table VII). substantial improvement in the cost effectiveness of
Table II of the supplementary material (see “Ar- palivizumab prophylaxis in preterm infants and chil-
dren with BPD, and a dominant outcome for prophy-ticlePlus” at www.adisonline.com/phe) shows the
laxis in CHD.results of various sensitivity analyses; the findings
The model was based on a number of conserva-correspond with those for prophylaxis in preterm
tive assumptions. It was assumed that palivizumabinfants and children with BPD.
only affects the occurrence of RSV hospitalisation
Figures 5 and 6 show the corresponding cost-
and does not influence the severity of the RSV
effectiveness acceptability curve and incremental infection. However, clinical trials suggest that
cost-effectiveness plane for children with CHD. palivizumab reduces the severity of the RSV infec-
These indicate a 0.98 probability that the ICER for tion. The IMpact[10]
and Feltes et al.[11]
trials showed
palivizumab was less than £25 000/QALY. an absolute reduction in probability of ICU admis-
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Cost-effectiveness threshold (£ × 1000)
Probabilityofbeingcosteffective
No prevention
Palivizumab
Fig. 5. Cost-effectiveness acceptability curve for palivizumab prophylaxis in congenital heart disease (discounting effectiveness and costs,
2003 values).

QALYS
Costs(£)
£25 000/QALY
−4000
−3000
−2000
−1000
0
1000
2000
3000
4000
5000
6000
7000
−0.4 −0.2 0 0.2 0.4 0.6 0.8 1.0 1.2
Fig. 6. Incremental cost-effectiveness plane for palivizumab prophylaxis in congenital heart disease (discounting effectiveness and costs,
2003 values).
sion for infants who received palivizumab (vs place- mortality. Over a wide range of varying values for
bo), which suggests a positive impact on the severity mortality, this evaluation was able to suggest cost
of the RSV infection, although the difference in the effectiveness for palivizumab under the current
Feltes et al.[11]
trial was not statistically significant. health economic standards.[40]
Figure 2 shows the
sensitivity of the ICER to mortality rates.The LOS for hospitalisation was also based on a
conservative approach. An analysis of the underly- Unlike previous studies, this study is not limited
ing IMpact[10] and Feltes et al.[11] study data showed to a short-term time horizon of one season and
that palivizumab patients had significantly shorter including only medical costs, but also includes long-
LOS than patients without prevention. Nevertheless, term discounted costs of asthma and productivity,
the analysis was based on the same LOS for which are relevant in the context of RSV infection, a
palivizumab and ‘no prophylaxis’ by pooling the disease with long-term clinical and economic conse-
data of both study arms. LOS may also be consid- quences. It also re-analyses unpublished trial data
ered a measure of the severity of RSV infection, for UK subjects to provide better estimates of re-
which confirms the previous hypothesis that source use. Another important difference is that this
palivizumab may reduce the severity of an RSV study uses the recommended cost-effectiveness out-
infection. The results of the model are also very come according to the NICE guidelines,[12]
i.e. cost/
sensitive to the vial use. The base-case analysis QALY, whereas other studies express cost effective-
reflects vial use according to the key clinical trials. ness in cost per avoided hospitalisation. This ICER
A scenario analysis examined the effect of vial measure is not a generally accepted cost-effective-
usage on costs and found that the base-case analysis ness outcome and no thresholds are available for this
was conservative. ICER measure. Consequently, the interpretation of
the cost effectiveness based on cost per hospitalisa-
The results of this study suggest a more favour-
tion avoided is subjective and should be considered
able cost-effectiveness outcome for palivizumab
cautiously.
than previously published studies in this area.[37-39]
The current study reflects a broader range of oppor- We made some additional assumptions that were
tunity costs by including long-term downstream conservative for the study drug. For example, only
consequences (e.g. asthma, productivity losses) and 2-year asthma cost data were available, whereas
also incorporating the effect of RSV infection on childhood asthma usually persists until 16–18 years

70 Nuijten et al.
3. Canfield S, Simoes E. Prevention of respiratory syncytial virus
of age. As with other models, we had to rely on
(RSV) infection: RSV immune globulin intravenous and
clinical trial data. Data on LOS were derived from palivizumab. Ped Ann 1999; 28: 507-14
4. Public Health Laboratory Service. Seasonal diseases: winterthe clinical trials, but we used a subset of UK
infections. Respiratory syncytial virus (RSV) infections [on-
patients in order to reflect treatment patterns in the line]. Available from URL: http://www.phls.co.uk/seasonal/
UK; treatment of RSV hospitalisation infection was rsv/RSV13.htm [Accessed 2005 Jun 15]
5. Clark S, Beresford M, Subhedar N, et al. Respiratory syncytialnot restricted by clinical protocol.
virus infection in high risk infants and the potential impact of
Given that there are some limitations to the mod- prophylaxis in a United Kingdom cohort. Arch Dis Child
2000; 83: 313-6elling technique, the advantages should also be
6. Sampalis JS. Morbidity and mortality after RSV-associatedpointed out. The modelling approach here allows for
hospitalizations among premature Canadian infants. J Pediatr
a difficult problem to be broken down into discreet 2003 Nov; 143 (5 Suppl.): S150-6
units that can be measured and analysed. A sensitivi- 7. Bont L, van Aalderen W, Kimpen J. Long term consequences of
respiratory syncytial virus (RSV) bronchiolitis. Paediatrty analysis was performed on the input variables of
Respir Rev 2000; 1: 221-7
the model where we felt that there was a question 8. Sigurs N, Bjarnason R, Sigurbergsson F, et al. Asthma and
immunoglobulin E antibodies after respiratory syncytial virusabout the results.
bronchiolitis: a prospective cohort study with matched con-
trols. Pediatrics 1995; 95: 500-5
Conclusion 9. American Academy of Pediatrics Committee on Infectious Dis-
eases and Committee of Fetus and Newborn. Prevention of
Taking into account the various considerations, respiratory syncytial virus infections: indications for the use of
palivizumab and update on the use of RSV-IGIV. Pediatricsthis analysis suggests that palivizumab in all its
1998 Nov; 102 (5): 1211-6
licensed indications may be a cost-effective prophy- 10. IMpact study group. Palivizumab, a humanised respiratory syn-
cytial virus monoclonal antibody, reduces hospitalisation fromlactic option against severe RSV infections in the
respiratory syncytial virus infection in high risk infants. Pedi-
UK versus no prophylaxis. Even given for one RSV atrics 1998; 102: 531-7
season, the positive clinical and economic benefits 11. Feltes TF, Cabalka AK, Meissner HC, et al. Palivizumab pro-
phylaxis reduces hospitalization due to respiratory syncytialmay persist over a lifetime analysis.
virus in young children with hemodynamically significant
congenital heart disease: Cardiac Synagis Study Group.
Acknowledgements J Pediatr 2003; 143: 532-40
12. NICE. Guidance for manufacturers and sponsors: technology
This study was funded by Abbott GmbH & Co. KG, appraisals No. 5, 2001 [online]. Available from URL: http://
Ludwigshafen, Germany. The authors are grateful to Prof. www.nice.org.uk/page.aspx?o=201974 [Accessed 2005 Jun
15]Adrian Bagust, Economics Division, University of Liverpool,
13. Personal Social Services Research Unit. Unit costs of health andUK, for his support and input in the development of this
social care 2003 [online]. Available from URL: http://www.pstudy. MEDTAP received funding from Abbott GmbH & Co.
ssru.ac.uk/pdf/uc2003/uc2003_inflation.pdf [Accessed 2005
KG, Ludwigshafen, Germany for the development of the Jun 1]
model and writing of the manuscript. Wolfgang Wittenberg is 14. Nuijten MJC. Data management in modelling studies: the selec-
an employee of Abbot GmbH and owns stock in the company. tion of data sources. PharmacoEconomics 1998; 3 (3): 305-16
Maximilian Lebmeier was a contractor for Abbott Germany. 15. Groothuis JR, Simoes EA, Levin MJ, et al. Prophylactic admin-
All the authors contributed significantly to the intellectual istration of respiratory syncytial virus immune globulin to
high-risk infants and young children: Respiratory Syncytialcontent of the paper.
Virus Immune Globulin Study Group. N Engl J Med 1993;
329: 1524-30
16. PREVENT Study Group. Reduction of RSV hospitalisationReferences
among premature infants and infants with bronchopulmonary1. Hall C. Respiratory syncytial virus. In: Feigin R, Cherry J,
dysplasia using respiratory syncytial virus immune globulineditors. Infectious diseases. Philadelphia (PA): W.B. Saun-
prophylaxis. Pediatrics 1997; 99: 93-9ders, 1998
2. Allport T, Davies E, Wells C, et al. Ribavarin and bronchiolitis: 17. Simoes EA, Sondheimer HM, Top FH, et al. Respiratory syncyt-
variation in use in the UK [letter]. Arch Dis Child 1997; 76: ial virus immune globulin for prophylaxis against respiratory
385 syncytial virus disease in infants and children with congenital

heart disease: Cardiac study group. J Pediatr 1998 Oct; 133 (4): www.dh.gov.uk/publicationsandstatistics/statistics/hospitalepi
492-9 sodestatistics/fs/en [Accessed 2005 Jun 15]
18. Joffe S, Ray GT, Escobar GJ, et al. Cost-effectiveness of respir- 31. Simoes EAF, Carbonell-Estrany X, Kimpen J, et al. Palavizum-
atory syncytial virus prophylaxis among preterm infants. Pedi- ab use decreases risk of recurrent wheezing in preterm chil-
atrics 1999 Sept; 104 (3 pt 9): 419-27 dren. Eur Respir J 2004; 24 Suppl. 48: 212s
19. MacDonald NE, Hall CB, Suffin SC, et al. Respiratory syncytial
32. International Labour Organization (ILO). Yearbook of Labour
viral infection in infants with congenital heart disease. N Engl
Statistics, 1998. Geneva: ILO, 1998
J Med 1982 Aug 12; 307 (7): 397-400
33. Organisation for Economic Co-operation and Development20. Rodriguez WJ, Parrott RH. Ribavirin aerosol treatment of seri-
(OECD). OECD employment outlook: 2004 edition [online].ous respiratory syncytial virus infection in infants. Infect Dis
Available from URL: http://puck.sourceoecd.org/vl=4300666/Clin North Am 1987 Jun; 1 (2): 425-39
cl=14/nw=1/rpsv/ij/oecdthemes/99980045/v2004n8/s1/p11
21. Wang EE, Law BJ, Stephens D. Pediatric investigators collabo- [Accessed 2006 Dec 1]
rative network on infections in Canada (PICNIC) prospective
34. The Chartered Institute of Public Finance and Accountancy. Thestudy of risk factors and outcomes in patients hospitalized with
health service database 2003 (financial year 2002-3). London:respiratory syncytial viral lower respiratory tract infection.
CIPFA/HFM, IPF, 2003J Pediatr 1995; 126: 212-9
22. Wang E, Tang NK. Immunoglobulin for preventing respiratory 35. Netten A, Curtis L. Unit costs of health and social care 2003.
syncytial virus infection (Cochrane Review). Available in The Canterbury, UK: Personal Social Services Research Unit, Uni-
Cochrane Library [database on disk and CD ROM]. Updated versity of Kent at Canterbury, 2000
quarterly. The Cochrane Collaboration; issue 2. Chichester
36. Devlin N, Parkin D. Does NICE have a cost-effectiveness(UK): Oxford Update Software, 2004
threshold and what other factors influence its decisions? A
23. Simoes E. Respiratory syncytial virus infection. Lancet 1999; binary choice analysis. Health Econ 2004; 13 (15): 437-52
354: 847-52
37. Stevens TP, Sinkin RA, Hall CB, et al. Respiratory syncytial
24. National statistics [online]. Available from URL: http://
virus and premature infants born at 32 weeks’ gestation or
www.statistics.gov.uk [Accessed 2005 Jun 1]
earlier: hospitalization and economic implications of prophy-
25. Wren C, O’Sullivan JJ. Survival with congenital heart disease laxis. Arch Pediatr Adolesc Med 2000 Jan; 154 (1): 55-61
and need for follow up in adult life. Heart 2001; 85 (4): 438-43
38. Marchetti A, Lau H, Magar R, et al. Impact of palivizumab on
26. Greenough A, Alexander J, Burgess S, et al. Health care utilisa- expected costs of respiratory syncytial virus infection in
tion of prematurely born, preschool children related to hos- preterm infants: potential for savings. Clin Ther 1999 Apr; 21
pitalisation for RSV infection. Arch Dis Child 2004; 89 (7): (4): 752-66
673-8
39. Rietveld E, Steyerberg E, Polder J, et al. Passive immunisation
27. Netten A, Curtis L. Unit costs of health and social care 2000.
against respiratory syncytial virus: a cost-effectiveness analy-
Canterbury, UK: Personal Social Services Research Unit, Uni-
sis [thesis]. Enschede: Print Partners Ipscamp, 2003: 59-69
versity of Kent at Canterbury, 2000
40. Rawlins MD, Culyer AJ. National Institute for Clinical Excel-28. Thomas M, Bedford-Russella A, Sharland M. Hospitalisation
lence and its value judgments. BMJ Jul 2004; 329: 224-7for RSV infection in ex-preterm infants: implications for use of
RSV immune globulin. Arch Dis Child 2000; 83: 122-7
29. Greenough A, Cox S, Alexander J, et al. Health care utilisation
Correspondence and offprints: Dr Mark J.C. Nuijten, Eras-of infants with chronic lung disease, related to hospitalisation
mus University, Postbus 1738, Rotterdam, 3000 DR, Thefor RSV Infection. Arch Dis Child 2001; 85: 463-8
Netherlands.30. Hospital Episode Statistics 1995-1999. London: Department of
E-mail: nuijten@bmg.eur.nlHealth, 2000 [online]. Available from URL: http://

RSV Prev CE Analysis

More Related Content

What's hot

Viewers also liked

Similar to RSV Prev CE Analysis

RSV Prev CE Analysis